Management of Wireless Local Area Networks by Artificial Neural Networks with Principal Components Analysis

Abstract

One of the main problems of a wireless local area networks (WLANs) management model is the difficulty for remote administrators to determine whether the wireless base station could provide proper connection services to users. SYSLOG (security issues in network event logging) records events occurring in wireless base stations and conveys the events back to administrators. This study employed back-propagation neural networks (BPNN) with principal components analysis (PCA) to analyze the SYSLOG data and the connection status between wireless base stations and users. The PCA technique was used to select essential SYSLOG data influencing connecting status; and the BPNN model was applied to categorize the connection status in terms of SYSLOG data. The simulation results indicated that the BPNN with PCA procedure is a feasible and promising way in the management of wireless local area networks.